Predictive analysis brief study of early-stage diabetes using multiple classifier models

Abstract

Diabetes Mellitus is one of the world's leading causes of mortality, with a worldwide death toll estimated to be in the millions. It is determined by the concentration of a sugar molecule in the blood, which is produced from glucose. Predicting the likelihood of contracting this illness may now be done using a plethora of methods. Data about diabetic patients must be comprehensive and accurate in order to accurately forecast the onset of the disease. In this paper, we discussed early-stage diabetes prediction using six algorithms. The algorithms are Gradient Boosting, ADA Boosting, XG Boosting, Neural Network, SVM, Random Forest, Stacking Neural Network, Stacking SVM, Stacking Random Forest. We also discussed briefly about the best algorithm among them with detailed accuracy by class and confusion matrix. By this study, we can predict early-stage diabetes disease more accurately.